Excellent catch -- thanks for letting us know. I've committed the fix and slated it over to the v1.8 branch for inclusion in 1.8.1.
On Apr 16, 2014, at 7:55 AM, Bastian Beischer <bastian.beischer_at_[hidden]> wrote:
> Dear all,
> In open-mpi version 1.8 the man page of MPI_Type_create_hindexed gives this declaration of the function:
> int MPI_Type_create_hindexed(int count, const int array_of_blocklengths,
> const int array_of_displacements, MPI_Datatype oldtype,
> MPI_Datatype *newtype)
> However I think the array_of_displacements should be of type MPI_Aint?
> This was relevant to me because I wanted to create a datatype with large offsets > INT_MAX and I didn't know I could use MPI_Type_create_hindexed until a colleague pointed me to the MPI standard which says that the displacements are of type MPI_Aint.
> Bastian Beischer
> RWTH Aachen University of Technology
> @RWTH Aachen
> Office: 28 C 203
> Phone: +49-241-80-27205
> E-mail: beischer_at_[hidden]
> Address: I. Physikalisches Institut B, Sommerfeldstr. 14, D-52074 Aachen
> Office: Bdg 32-4-B12
> Phone: +41-22-76-75750
> E-mail: bastian.beischer_at_[hidden]
> Address: CERN, CH-1211 Geneve 23
> devel mailing list
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/devel
> Link to this post: http://www.open-mpi.org/community/lists/devel/2014/04/14540.php
For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/